Meeting Agenda

1. Arizona Health Care Cost Containment SystemDRG-Based Inpatient Hospital Payment SystemDRG WorkgroupDecember 6, 2012

2. Meeting Agenda

3. Evaluation of APR-DRGs

4. Evaluation of APR-DRGs Source: Quinn, K, Courts, C. Sound Practices in Medicaid Payment for Hospital Care. CHCS: November 2010, updated with current information by Navigant.

5. Evaluation of APR-DRGs Source: Quinn, K, Courts, C. Sound Practices in Medicaid Payment for Hospital Care. CHCS: November 2010, updated with current information by Navigant.

6. Evaluation of APR-DRGs • Validation of 3M national relative weights • Calculated Arizona Medicaid-specific APR-DRG relative weights to compare to 3M’s national weights • Arizona weights based on average cost per discharge by DRG from model, adjusted for wage index • Able to calculate stable relative weights for 604 out of total 1,256 DRG/SOI combinations • DRGs had to have at least 10 cases • DRGs had to have enough cases to pass stability test • SOI weights had to be ascending across same DRG

7. Evaluation of APR-DRGs • Weight Correlation Analysis

8. Evaluation of APR-DRGs • Weight Correlation Analysis

9. Preliminary Baseline Model

10. Preliminary Baseline Model • Preliminary Models for Discussion Purposes Only • Note that at this time, no final decisions have been made or proposed by AHCCCS • These preliminary analyses have been prepared by Navigant for discussion purposes only, and do not necessarily reflect recommendations by AHCCCS or Navigant • Actual new DRG payment system results may be significantly different from these preliminary projections due to payment methodology changes and future changes in patient volume and case mix

11. Preliminary Baseline Model • Baseline Model Assumptions (refined since September meeting) • APR-DRG version 30 model using 3M national weights (adjusted by a factor of 0.75725 to achieve an average Arizona case mix index of 1.0) • Statewide standardized DRG base rate of $4,917.77, with labor portion adjusted by FFY 2013 Medicare wage index • Medicare-style outlier and transfer payment policy • Outlier fixed-loss threshold of $57,500 set to achieve approximately 6% outlier payments • No policy adjustors for select services or providers

12. Preliminary Baseline Model • Baseline Model Assumptions – Claims/Encounter Data • Model based on FFY 2010 Arizona Medicaid FFS claims data and MCO encounter data collected from AHCCCS (dates of service from 10/1/2009 through 9/30/2010) • Includes in-state general acute providers, CAHs and select out-of-state providers • Excludes Medicare dual-eligibiles, federally-funded FFS claims, 638/IHS providers, and cases with “ungroupable” APR-DRG classifications • Excludes freestanding psychiatric provider cases and Maricopa psychiatric cases with transitional stays

13. Preliminary Baseline Model • Baseline Model Assumptions – Funding Pool • Model funding pool for new DRG system based on combined reported FFS claim and MCO encounter data reported payments, with adjustments for rate reductions • 0.9025 factor applied to reported payments to reflect 5% rate reductions that occurred on 10/1/2010 and 10/1/2011 • DRG funding pool does not include static payments

14. Preliminary Baseline Model • Baseline Model Assumptions – Funding Pool • Model designed such that simulated aggregate payments under new DRG system are equal to total DRG funding pool • For evaluation purposes, 2012 static payments have been allocated to the individual claims/encounter data in the analytical dataset • Allocation based on relative charges (excluding DSH and Safety Net Pool payments) • Model assumes no changes to static payments for each hospital under new system

15. Preliminary Baseline Model • Baseline Model Assumptions – Estimated Costs • Costs estimated at a detail line level by applying each provider’s cost center-specific cost-to-charge ratios (CCRs) to ancillary revenue code charges and cost per diems to routine revenue code days • CCRs and cost per diems calculated from hospital Medicare cost report data extracted from the HCRIS dataset with fiscal reporting periods matching the FFY 2010 claim/encounter dataset • Estimated costs inflated from FFY 2010 to FFY 2014 by a factor of 11.6%, based on changes in CMS input price index levels

16. Preliminary Baseline Model • Preliminary Baseline Model Results

17. Outlier Payment Model Comparison

18. Outlier Payment Model Comparison

19. Policy Adjustors Impact

20. Policy Adjustors Impact • Policy Adjustor Model Assumptions • Uses baseline model as starting point • Same claims/encounter data, funding pool and estimated costs • Same APR-DRGs, relative weights, transplant and outlier payment methodologies • Applied policy adjustors to DRG base payments for key Medicaid services • Statewide standardized DRG base rate of $4,387.28, with labor portion adjusted by FFY 2013 Medicare wage index

21. Policy Adjustors Impact • Policy Adjustors by Service • Preliminary model policy adjustors applied to DRG base payments: • Normal newborn DRGs: 1.40 factor • Neonate DRGs: 1.15 factor • Obstetric DRGs: 1.45 factor • Psychiatric/Rehabilitation DRGs: 1.45 factor • Other pediatric cases (age 18 and under): 1.15 factor • Policy adjustors set for each service to achieve same pay-to-cost ratio as statewide average (including allocated static payments) • Exception: Psychiatric/Rehabilitation policy adjustor set to achieve current system spending

22. Policy Adjustors Impact • Preliminary Policy Adjustor Model Results

23. Policy Adjustors Impact • Model comparison – Estimated Pay-to-Cost Ratio

24. Policy Adjustors Impact • Model comparison – Estimated Payment Change

25. Coding Documentation and Improvement Adjustment

26. Coding and Documentation Improvement Adjustment • Why do we need an adjustment? • Coding and documentation improvements are a necessary and appropriate response by providers to the requirements under the APR-DRG model. • Because the same level of coding rigor was not required for payment purposes under the legacy per diem model, we assume that case mix in our simulation models is understated. • We expect that case mix will increase in future periods, beyond actual increases in patient acuity. • Increases cannot be predicted with precision – and may be significant.

27. Coding and Documentation Improvement Adjustment Coding requirements are significantly different for APR-DRGs, even when compared to the requirements under the Medicare MS-DRG model.

28. Coding and Documentation Improvement Adjustment • Calculated Model Case Mix • Calculated case mix for two large urban trauma providers with similar Medicaid volume:

29. Coding and Documentation Improvement Adjustment Examples of Actual Case Mix Increases from DRG Grouper Change • In October 2007, CMS in its Medicare Inpatient Prospective Payment System (IPPS) replaced its CMS-DRG grouper with the MS-DRG grouper • CMS subsequently estimated that the extent of case mix increase from coding improvements above real case mix for FFY 2008-2009 was 5.8% • Medicare inpatient Documentation and Coding Adjustment preemptively reduces rates; 2.0% FFY 2012 and 1.9% in FFY 2013 • In July 2010, the Pennsylvania Department of Public Welfare (DPW) in its Medicaid IPPS replaced its CMS-DRG grouper with the APR-DRG grouper • DPW subsequently estimated that total case mix increases for SFY 2011 was 12.1%

30. Coding and Documentation Improvement Adjustment Higher Rate of Paid Casemix Increases Return to Pre-Implementation Levels Bump from CDI System Implementation Paid Casemix Rate of Increase Without APR-DRG Implementation Lower

31. Coding and Documentation Improvement Adjustment 1.06 Aggregate CMI

32. Transition Period

33. Transition Period • Example Transition Period • Payments are made through DRG methodology • Transition is created through adjustment to hospital base rates • Limits individual hospital’s payment change to: • +/- 10% of simulated payment change amount in year 1 • +/- 20% of simulated payment change amount in year 2 • +/- 40% of simulated payment change amount in year 3 • Rebase using cases paid under APR-DRGs and coded under ICD-10 in year 4 • Will allow hospitals time to adjust, improve efficiency, and reduce cost growth • Actual transition period may differ from example

34. Questions and Discussion

35. Questions and Discussion Questions and comments may be addressed to Jean Ellen Schulik at JeanEllen.Schulik@azahcccs.gov (602) 417-4335 DRG Project Website: http://www.azahcccs.gov/commercial/ProviderBilling/DRGBasedPayments.aspx